On TikTok, some viewers recently saw an ad in which MrBeast offered a deal that seemed too good to be true. It was.
In the spot, MrBeast (whose real name is Jimmy Donaldson) promised to give away 10,000 iPhone 15 Pro devices for just $2 apiece. But days after the promotion went live, the real Jimmy Donaldson took to X to tell viewers that he did not film the ad himself. The character seen in the video was nothing more than an AI-generated deepfake designed to resemble the 25-year-old creator.
“Lots of people are getting this deepfake scam ad of me…are social media platforms ready to handle the rise of AI deepfakes?” Donaldson wrote. “This is a serious problem.”
Subscribe for daily Tubefilter Top Stories
Lots of people are getting this deepfake scam ad of me… are social media platforms ready to handle the rise of AI deepfakes? This is a serious problem pic.twitter.com/llkhxswQSw
— MrBeast (@MrBeast) October 3, 2023
Though Donaldson is known for his generous giveaways, several TikTok viewers noticed the scammy nature of the ad. The deepfake version of MrBeast didn’t sound quite right, and its lips didn’t line up with its voice. “Only boomers would fall for this shit,” wrote one commenter.
Deepfakes of public figures are banned on TikTok. The ad was removed a few hours after it was uploaded, but onlookers are still wondering how it slipped past the moderation team in the first place. TechCrunch suggested that the AI-generated character may have fooled the AI-assisted technology TikTok uses to assess ads before they go live.
TikTok is not the only platform struggling to contain deepfakes. Despite widespread bans, bad actors continue to get away with ersatz representations of celebrities. Aside from Donaldson, other individuals who have recently been impersonated by deepfakes include Gayle King and Tom Hanks.
So what can moderation teams do moving forward? In a response to MrBeast’s TikTok callout, fellow creator Kwebbelkop (real name Jordi van den Bussche) laid out a few ideas. He argued that facial recognition tools, hidden watermarks, and stronger punishments for deepfakers could help solve the problem. “Definitely a big issue!” he wrote. “Fixes aren’t too complicated.”
